Genetic Mechanism of Structurally Controlled Dolomites Derived from Seawater-Hydrothermal Mixed Fluids—A Case Study from Middle Permian, Central Sichuan Basin, South China

: Dolomite bodies in the Middle Permian of the central Sichuan Basin have been reported as favorable natural gas reservoirs. The Middle Permian dolomite consists of three types of recrystallized dolomite (Rd1, Rd2, and Rd3) and one type of dolomite cement (Sd). Rd1 might be formed as the primary mineral along the calcite in the original sea-water. Its δ 13 C value and 87 Sr/ 86 Sr ratio, consistent with those of marine limestone and Permian seawater, support that the dolomitizing ﬂuid for Rd1 was Permian seawater preserved in the strata. Rd2 consists of ﬁne to medium (50 µ m to 250 µ m) and planar to curved crystals. Geochemical indicators (slightly high 87 Sr/ 86 Sr ratio, similar rare earth element patterns, negative δ 18 O, slightly high salinity) conﬁrm that the dolomitizing ﬂuid of Rd2 was mainly Permian seawater during shallow burial, with a small number of hydrothermal ﬂuids. Rd3 and Sd are featured by very large (>250 µ m), curved crystals, and high-temperature, high-salinity, and obviously positive Eu anomalies, suggesting that their diagenetic ﬂuids were mainly hydrothermal ﬂuids from deep. Additionally, inherited carbon sources and the 87 Sr/ 86 Sr ratios of some samples fall within the range of Permian seawater distribution, conﬁrming the contribution of Permian seawater. ELIP activity caused the formation of this dolomite through the mixing of seawater and hydrothermal ﬂuids. The main ﬂuid circulation channels were activated basement faults, epigenetic karst pores, and shallowly buried high-permeability strata. During the peak period of ELIP activity, the continuous upwelling of deep hydrothermal ﬂuids led to the continuous formation of Rd2, Rd3, and Sd. The dolomitization ﬂuid of Rd2 was mainly composed of seawater and featured a certain lateral extension, which was away from faults. Rd3 and Sd are mainly distributed along the fault system, and excessive dolomitization caused by the hydrothermal activity, to some extent, inhibited the lateral movement of hydrothermal ﬂuids. This study provides a good example for exploring the genetic mechanism and distribution pattern of structurally controlled dolomites under a volcanic activity background.


Introduction
Hydrothermal dolomitization, which is controlled by basement faults, is widely discussed as a distinctive type of dolomitization on a global scale [1][2][3][4]. Generally, this type of dolomitization typically occurs in extensional tectonic geological environments and can result in the formation of exceptional dolomite oil and gas reservoirs [5][6][7]. Deep hydrothermal fluids, which are associated with tectonic or volcanic activity, are widely recognized as playing a pivotal role in the process of hydrothermal dolomitization, as evidenced by numerous studies in locations such as the Western Canada sedimentary basin, Sichuan Basin, and the Tarim Basin in China [3,[8][9][10][11]. However, emerging research suggests that seawater also exerts a significant influence on the hydrothermal dolomitization process, potentially serving as the primary source of Mg 2+ [7,12,13]. The accurate identification of dolomitizing fluids carries important implications for comprehending the mechanisms of dolomitization and predicting the distribution patterns of dolomite reservoirs.
The dolomite layers within the Middle Permian strata of the Sichuan Basin have been repeatedly reported as excellent hydrocarbon reservoirs [12,[14][15][16][17][18][19]. In the central Sichuan Basin, the distribution of Middle Permian dolomite, to some extent, exhibits characteristics that are controlled by basement faults [12,15,20]. Accordingly, most studies suggest that the genesis of the dolomite is associated with deep hydrothermal activity [18][19][20][21]. Additionally, the magmatic activity of the Emeishan Large Igneous Province (ELIP) is believed to have created anomalous hydrothermal fluids and the associated thermal effects across the basin [22][23][24]. However, some studies suggested that the formation processes of Middle Permian dolomites were extremely complex and were the result of multiple stages of superposition [9,16]. Currently, these researches are not clear enough on tracing the sources of diagenetic fluids for each stage, which hinders the understanding of dolomite formation mechanisms. The dolomite bodies in the Middle Permian exhibit a certain degree of heterogeneity in the vertical direction, and the lack of discussion on this phenomenon also limits the prediction of its reservoir distribution.
This study focuses on the typical outcrops and drill cores in the central Sichuan Basin. Microscopic petrology was utilized to characterize different lithofacies types of the dolomite and determine the diagenetic sequence. Geochemical analyses, including C, O, and Sr isotopes and trace elements, were conducted to analyze the geochemical properties of the diagenetic fluids. Subsequently, fluid inclusion analysis was performed to constrain the temperature and salinity of the diagenetic fluids. Our study has effectively revealed the origins of the various stages of dolomitizing fluids in the central Sichuan Basin. We have also discussed the hydrological cycle and water-rock interaction processes, which have explained the spatiotemporal distribution of different types of dolomites. The findings of this study are significant for understanding the formation mechanisms of shallow-buried dolomites affected by igneous activity and have potential implications for the prediction of related dolomite reservoirs.

Geological Setting
The Sichuan Basin, located in southwestern China (Figure 1a), is a large hydrocarbonbearing basin. It is a secondary tectonic unit on the western margin of the Yangtze Platform and has a rhombus-like shape, bordered by several major tectonic belts [25,26]. The Middle Permian strata in the study area comprise the Liangshan Formation (P 2 l), Qixia Formation (P 2 q), and Maokou Formation (P 2 m), in ascending order. The Liangshan Formation was deposited in a coastal environment consisting mainly of coal-bearing mudstone, carbonaceous shale, and sandstone. The Qixia and Maokou Formations were deposited in a stable tropical to subtropical carbonate platform environment, consisting of thick and homogeneous carbonate deposits. The Qixia Formation comprises two members (Qi1 and Qi2), which were deposited in relatively shallower water than the Maokou Formation, particularly in the central Sichuan region, where bioclastic and intraclastic shoals were well developed. The Maokou Formation consists of four members (Mao1 to Mao4). During the early depositional stage of the Maokou Formation, a large-scale transgression occurred in the study area, leading to a relatively deep-water sedimentary environment throughout the area, with the Mao1 dominated by rhythmic layers of argillaceous limestone and marl. This was followed by a sustained slow regression, with gradually shallowing water depth, and the Mao2 and Mao3 were mainly composed of bioclastic limestone and grainstone. Due to the Dongwu Movement-induced uplift, the top of the Maokou Formation in the study area was extensively eroded [27][28][29]. The development of dolomite in both vertical and horizontal directions shows strong heterogeneity in the central Sichuan Basin. In the west of the study area (MX42, MX108, GC2, GT2), dolomite mainly develops in Qi2, while in the southeast of the study area (Tl6, XYD, GZS, HLC), dolomite mainly develops in Mao3 and shows the characteristics of distribution along basement faults (Figure 1b) [15,19].
The Sichuan Basin is intersected by NW-SE and NE-SW-oriented basement faults ( Figure 1b) [25,30]. These faults, which are confirmed to have been in an extensional state during the mid-to-late Permian, display features of cutting through the Middle Permian strata on seismic profiles [31,32]. The uplift and intense eruption of the ELIP is considered to have caused the tectonic extension background of the mid-late Permian, along with triggering a pronounced anomalous heat effect, manifested by a sustained increase in paleoheat flow values [22,24]. This anomaly is regional in scale and is not limited to the vicinity of the eruption center. In addition, the abundant ELIP basalts found in drill cores and outcrops in the central Sichuan Basin suggest the occurrence of fault-related hydrothermal activity and anomalous heat effects in this area [33]. During the early depositional stage of the Maokou Formation, a large-scale transgression occurred in the study area, leading to a relatively deep-water sedimentary environment throughout the area, with the Mao1 dominated by rhythmic layers of argillaceous limestone and marl. This was followed by a sustained slow regression, with gradually shallowing water depth, and the Mao2 and Mao3 were mainly composed of bioclastic limestone and grainstone. Due to the Dongwu Movement-induced uplift, the top of the Maokou Formation in the study area was extensively eroded [27][28][29]. The development of dolomite in both vertical and horizontal directions shows strong heterogeneity in the central Sichuan Basin. In the west of the study area (MX42, MX108, GC2, GT2), dolomite mainly develops in Qi2, while in the southeast of the study area (Tl6, XYD, GZS, HLC), dolomite mainly develops in Mao3 and shows the characteristics of distribution along basement faults (Figure 1b) [15,19]. The Sichuan Basin is intersected by NW-SE and NE-SW-oriented basement faults ( Figure 1b) [25,30]. These faults, which are confirmed to have been in an extensional state during the mid-to-late Permian, display features of cutting through the Middle Permian strata on seismic profiles [31,32]. The uplift and intense eruption of the ELIP is considered to have caused the tectonic extension background of the mid-late Permian, along with triggering a pronounced anomalous heat effect, manifested by a sustained increase in paleo-heat flow values [22,24]. This anomaly is regional in scale and is not limited to the vicinity of the eruption center. In addition, the abundant ELIP basalts found in drill cores and outcrops in the central Sichuan Basin suggest the occurrence of fault-related hydrothermal activity and anomalous heat effects in this area [33].

Materials and Methods
Samples of limestones and dolomites were collected from 7 drill cores and 5 outcrops in the Middle Permian in the central Sichuan Basin (Figure 1b). After detailed observations of outcrops and drill core samples, microscopic petrographic analysis was conducted at the Department of Geological Engineering, Chongqing Jiaotong University, using a polarizing microscope (NIKON LV100N). All thin sections were stained with Alizarin Red to distinguish dolomite from calcite [34]. Subsequently, 1 cm 3 blocks of the samples were cut and observed under a scanning electron microscope (SEM) using an FEI-QUANTA650FEG at the School of Earth and Space Sciences at Peking University. Moreover, energy-dispersive spectroscopy was employed to identify minerals within the carbonate rocks.
Geochemical analyses were conducted at the Key Laboratory of Orogenic Belts and Crustal Evolution and the School of Archaeology and Museology at Peking University. Mineral powders were obtained from the rocks using a micro-drilling technique, with

Materials and Methods
Samples of limestones and dolomites were collected from 7 drill cores and 5 outcrops in the Middle Permian in the central Sichuan Basin (Figure 1b). After detailed observations of outcrops and drill core samples, microscopic petrographic analysis was conducted at the Department of Geological Engineering, Chongqing Jiaotong University, using a polarizing microscope (NIKON LV100N). All thin sections were stained with Alizarin Red to distinguish dolomite from calcite [34]. Subsequently, 1 cm 3 blocks of the samples were cut and observed under a scanning electron microscope (SEM) using an FEI-QUANTA650FEG at the School of Earth and Space Sciences at Peking University. Moreover, energy-dispersive spectroscopy was employed to identify minerals within the carbonate rocks.
Geochemical analyses were conducted at the Key Laboratory of Orogenic Belts and Crustal Evolution and the School of Archaeology and Museology at Peking University. Mineral powders were obtained from the rocks using a micro-drilling technique, with efforts made to minimize potential interferences and contamination among different minerals. For trace element analyses, including rare earth elements (REEs), 80 mg of each sample powder was dissolved in 1 mL of HNO 3 (1 + 1) and evaporated. The residue was re-dissolved in 1.42 g/mL of HNO 3 and heated for 3 h and then was diluted with ultrapure water and tested for trace element compositions using an inductively coupled plasma mass spectrometer (ICP-MS). The post-Archean average shale (PAAS) was employed to standardize rare earth element concentrations [35]. The 87 Sr/ 86 Sr ratios were analyzed using a TRITON mass spectrometer. In the pre-treatment stage, 150 mg of sample powder was dissolved in an HCl solution (2.5 mol/L) and centrifuged for 8 min. The clear liquid in the upper layer of the centrifuged tube was then added to a cation exchange column and leached with HCl to extract pure Sr. The NBS987 standard was used to calibrate the obtained 87 Sr/ 86 Sr ratios, with an average testing precision of ±1.0 × 10 −5 (2δ). The carbon and oxygen isotope analyses were conducted using the IsoPrime 100 instrument. To prepare the samples for analysis, 200 mg of powder was dissolved in 99% H 3 PO 4 . The carbon and oxygen isotope values were calibrated against the IAEA CO-8 calcite standard (VPDB). The precision of the analysis was ±0.1‰. The fluid inclusion analysis was carried out at the School of Earth Sciences, Yunnan University. The homogenization temperature (T h ) and final melting temperature (T m-ice ) were determined by microthermometry of 100 µm thick, double-sided polished thin sections using a Linkam THMGS600. A cyclic heating method was applied to measure Tm-ice to achieve higher accuracy. Eighty-three fluid inclusions were analyzed in this study, and the salinity was calculated using the calculation formula established by Steele-MacInnis [36,37].

Microscopic Petrography
The Middle Permian dolomites are composed of three types of matrix-replaced dolomites (Rd1, Rd2, and Rd3) and dolomite cement (Sd) in the central Sichuan Basin. Rd1, which develops in an incompletely dolomitized state within the host limestone (Figure 3), is present in low abundance (<10%) in the dolomite formation and is composed of finegrained dolomite particles with planar crystals (Figure 3). Dolomitization preferentially occurred around stylolite and within the mudstone matrix in the host limestone, leading to the formation of Rd1 (Figure 3a,b). Rd1 displays rhombic shapes with fog center-bright edge structure (Figure 3c

Microscopic Petrography
The Middle Permian dolomites are composed of three types of matrix-replaced dolomites (Rd1, Rd2, and Rd3) and dolomite cement (Sd) in the central Sichuan Basin. Rd1, which develops in an incompletely dolomitized state within the host limestone ( Figure 3), is present in low abundance (<10%) in the dolomite formation and is composed of finegrained dolomite particles with planar crystals (Figure 3). Dolomitization preferentially occurred around stylolite and within the mudstone matrix in the host limestone, leading to the formation of Rd1 (Figure 3a,b). Rd1 displays rhombic shapes with fog center-bright edge structure (Figure 3c,d). Under cathodoluminescence (CL), Rd1 displays a dark red luminescence, while the host limestone shows almost no luminescence (Figure 3e,f).
The content of Rd2 in the dolostone unit is approximately 40%-50%, consisting of fine to medium-sized dolomite crystals (50~250 µm) with planar crystals (Figure 4). The dolomite crystals in Rd2 generally lack the original limestone textures, showing interlocking mosaic contacts with bright rimmed edges (Figure 4a). Rd2 displays the development of a reticulated crack system and brighter red luminescence than Rd1 under cathodoluminescence (CL), while the cement in the crack system emits similar bright red luminescence (Figure 4b,c). Cracks filled with both calcite and quartz are developed in Rd2 ( Figure  4d). Cracks filled with Sd, which cuts across Rd2, display a relatively bright orange-red luminescence in both minerals under CL (Figure 4e,f).
Rd3 constitutes 30%-40% of the volume of the dolomite, consisting of medium-to coarse-sized (250 µm-2 mm) crystals with curved surfaces (Figure 5a). Saddle dolomite (Sd) accounts for approximately 10%-20% of the volume of the dolomite and fills the fractures and pores of the matrix-replaced dolomite as cement ( Figure 5b). Typically, some Sd exhibit large crystals, which can exceed 2 mm in size, and display strong undulatory   The content of Rd2 in the dolostone unit is approximately 40%-50%, consisting of fine to medium-sized dolomite crystals (50~250 µm) with planar crystals (Figure 4). The dolomite crystals in Rd2 generally lack the original limestone textures, showing interlocking mosaic contacts with bright rimmed edges ( Figure 4a). Rd2 displays the development of a reticulated crack system and brighter red luminescence than Rd1 under cathodoluminescence (CL), while the cement in the crack system emits similar bright red luminescence (Figure 4b,c). Cracks filled with both calcite and quartz are developed in Rd2 (Figure 4d). Cracks filled with Sd, which cuts across Rd2, display a relatively bright orange-red luminescence in both minerals under CL (Figure 4e,f).
Rd3 constitutes 30%-40% of the volume of the dolomite, consisting of medium-to coarse-sized (250 µm-2 mm) crystals with curved surfaces (Figure 5a     photograph of (c) shows that Sd emits brighter orange-red light than Qz, which emits dimmer light; (e) The cross-polarized light photograph of Sd and Cc reveals strong wave extinction of Sd, GZS, P 2 m 3 ; (f) The cathodoluminescence photograph of (e) shows Sd emitting intense orange-red light, while Cc emits relatively dimmer orange-red light; (g) Rd2, Rd3, and Sd are sequentially filled from the host rock to the fracture, XYD, P 2 m 3 ; (h) The cathodoluminescence photograph of (g) shows that Rd2 and Rd3 crystals emit dim red light, while Sd exhibits an interlocking texture emitting bright red light at the rim.

Microthermometry and Salinity
Gas-liquid two-phase fluid inclusions in the form of isolated and fluid inclusion assemblage (FIA) were analyzed in Rd2, Rd3, and Sd (Figures 7 and 8

Microthermometry and Salinity
Gas-liquid two-phase fluid inclusions in the form of isolated and fluid inclusion assemblage (FIA) were analyzed in Rd2, Rd3, and Sd (Figures 7 and 8, Table 1). In Rd2, T h values range from 83 to 134 • C (average 109.4 • C), and salinity values calculated using T m-ice range from 4.3 wt% to 7.3 wt% (average 5.7 wt%). The temperatures and salinities of fluid inclusions (FI) in Rd3 are higher than those in Rd2, ranging from 104 to 200 • C (average 140 • C) and 5.1 wt% to 22.1 wt% (average 15.1 wt%). FIs in Sd feature the highest temperature and salinity values, ranging from 118 to 208 • C (average 160.8 • C) and 8.4 wt% to 20.1 wt% (average 16.1 wt%).          Table 2).

Trace and Rare Earth Elements
The REE patterns of limestone and Rd2 are relatively flat with slight LREE depletion, i.e., the lowest LREE/HREE ratios of 0.66 and 0.50, respectively (Figure 9a,b, Table 3). Rd3 and Sd exhibit average LREE/HREE ratios of 1.17 and 1.97, respectively, showing slight LREE enrichment (Figure 9c,d, Table 3). Notably, the presence of significant Ce anomalies was not observed in all of the examined samples (Figure 10a). In Rd3 and Sd samples, the Eu contents exhibit prominent positive anomalies on the REE patterns (Figure 9c,d). The average Eu/Eu* values of Rd3 and Sd are 6.6 and 11.7 (Figure 11b-d). Limestone shows the highest Sr content (average 724 ppm) and the lowest Mn content (average 64.5 ppm) (Figure 11c, Table 3). The average Sr and Mn contents of Rd2 are 197.1 ppm and 165 ppm (Figure 11c, Table 3). The Sr contents of Rd3 and Sd are 131.3 ppm and 228.4 ppm, and the Mn contents are 139.7 ppm and 170.3 ppm (Figure 11c, Table 3).
The 87 Sr/ 86 Sr ratios of limestone and Rd1 are the lowest among all the samples, ranging from 0.707219 to 0.708009 (average 0.707526) and 0.706914 to 0.707723 (average 0.707293), falling within the distribution range of the Permian seawater (0.7068-0.7081) [38,39]. The 87 Sr/ 86 Sr ratios of Rd2 are slightly higher than those of Rd1, ranging from 0.7074 to 0.708328 (average 0.70776), with the ratios of a few samples higher than the Permian seawater range. The 87 Sr/ 86 Sr ratios of Rd3 and Sd samples are the highest among all the samples, ranging from 0.7086 to 0.709234 (average 0.7085) and 0.707457 to 0.709453 (average 0.708426), respectively, in which the ratios most of the samples are higher than the Permian seawater distribution range (Figure 12).
was not observed in all of the examined samples (Figure 10a). In Rd3 and Sd samples, the Eu contents exhibit prominent positive anomalies on the REE patterns (Figure 9c,d). The average Eu/Eu* values of Rd3 and Sd are 6.6 and 11.7 (Figure 11b-d). Limestone shows the highest Sr content (average 724 ppm) and the lowest Mn content (average 64.5 ppm) (Figure 11c, Table 3). The average Sr and Mn contents of Rd2 are 197.1 ppm and 165 ppm (Figure 11c, Table 3). The Sr contents of Rd3 and Sd are 131.3 ppm and 228.4 ppm, and the Mn contents are 139.7 ppm and 170.3 ppm (Figure 11c, Table 3).          [38,39]. The 87 Sr/ 86 Sr ratios of Rd2 are slightly higher than those of Rd1, ranging from 0.7074 to 0.708328 (average 0.70776), with the ratios of a few samples higher than the Permian seawater range. The 87 Sr/ 86 Sr ratios of Rd3 and Sd samples are the highest among all the samples, ranging from 0.7086 to 0.709234 (average 0.7085) and 0.707457 to 0.709453 (average 0.708426), respectively, in which the ratios most of the samples are higher than the Permian seawater distribution range ( Figure 12).

Diagenetic Sequence
The diagenetic sequence of the three matrix-replaced dolomites (Rd1, Rd2, Rd3) and types of cement (Sd, Cc, Qz) is mainly determined by their paragenesis, crystal characteristics, and inter-relationships.
Rd1 primarily occurs in an incomplete dolomitization form in host limestones ( Figure 3) and replaced the mud-crystal matrix preferentially. This is consistent with the formation of dolomite seed crystals during the initial stage of dolomitization of mud [40]. Thus, it is indicated that Rd1 formed during the initial stages of dolomitization. Additionally, some Rd1 samples exhibit fine-to-medium crystal characteristics, which suggest that they have undergone recrystallization of fine-to-medium crystal Rd1. Similarly, remnants of Rd1 are visible in the dirtier central part of Rd2, also suggesting the recrystallization of Rd1 to Rd2 (Figure 4a). Furthermore, Rd2 with fine to medium planar to curved crystals evolved into Rd3 with non-planar medium to coarse crystals (Figure 5a), indicating that Rd3 was formed after Rd2.
The non-planar morphology and strong undulatory extinction of Rd3 suggest that it was formed at a higher temperature exceeding the critical roughening temperature (CRT) [41][42][43]. Sd commonly occurs as rim cement in the pores/fractures of Rd2 and Rd3, indicating its later formation than Rd2 and Rd3 (Figure 2g-i). Therefore, the continuous growth of Rd1, Rd2, Rd3, and Sd from matrix to pores (Figure 5g,h) suggests a recrystallization process of Rd1 to Rd3 and a precipitation process of Sd [1,44,45]. Furthermore, Rd2 was formed during the shallow burial stage, evidenced by the Rd2 crystals crosscut by stylolites (Figure 4b) [46,47].
Cc and Qz filling in the fractures, especially in Sd, indicate that their formation occurred later than all phases of dolomitization (Figures 2e and 5c,d). In addition, hydrothermal minerals, including fluorite, pyrite, and siderite, are observed to fill the voids in the pore space of Rd2, Rd3, and Sd ( Figure 6). Therefore, the mineral formation sequence in the middle Permian dolomite of the study area can be ordered from early to late as follows: Rd1, Rd2, Rd3, Sd, Cc, Qz, and various hydrothermal minerals.

Sources of Dolomitizing Fluids
Several petrological and geochemical indicators suggest that the dolomitization fluids responsible for the formation of Rd1 were derived from preserved Permian seawater in the strata. The 87 Sr/ 86 Sr ratios of Rd1, which fall within the distribution range of the contemporaneous seawater 87 Sr/ 86 Sr ratios during the Permian (0.7068-0.7081), indicating that the dolomitization fluids of Rd1 were primarily derived from coeval seawater ( Figure 12). Therefore, it is concluded that Rd1 was formed from dolomitizing fluids derived from preserved Permian seawater in the sedimentary sequence.
Multiple pieces of evidence suggest that the dolomitizing fluid responsible for the formation of Rd2 was mainly derived from Permian seawater, and it may have incorporated a minor component of externally-derived hydrothermal fluids. The δ 13 C values of Rd2 exhibit consistency with those of Rd1 and the host limestones, indicating the inheritance of their carbon sources. Moreover, Rd2 and limestones exhibit similar REE distribution patterns characterized by relatively flat curves (Figure 9a,b), suggesting a consanguineous nature of their mineralizing fluids [48,49]. The lack of obvious Eu and Ce anomalies further suggests that the diagenetic fluids did not experience a deeply buried reducing environment [50,51]. However, some Rd2 samples exhibit 87 Sr/ 86 Sr ratios higher than the range of the Permian seawater ( Figure 12), indicating that some external fluids may have participated in the water-rock process. The formation of Rd2 is considered to be in shallow burial environments (<500 m) due to the crosscut of stylolites through Rd2 (Figure 4a). Similarly, the Sr contents in Rd2 are significantly lower than those in the limestone samples but markedly higher than those commonly observed in deeply buried limestones [52,53], indicating shallow burial diagenesis. This is because Sr is easily released from the crystal lattice due to its chemical bonding type and electronegativity characteristics that are similar to Ca but different from Mg during the diagenetic process of progressive burial [54].
The fluid inclusions in the clear rims of Rd2 exhibit a uniform temperature range of 83 to 134 • C (average 109.4 • C), which is significantly higher than the temperature calculated based on the maximum shallow-burial depth (i.e., 500 m), confirming that the formation temperature of its diagenetic fluid was extremely high. In addition, the more negative δ 18 O values than those of Rd1 ( Figure 11) also indicate that the formation temperature of Rd2 was higher than that of Rd1. It is worth noting that the salinity range of the Rd2 inclusions, calculated based on T m-ice , is slightly higher than that of the Permian seawater (4.4 wt%, [55,56]), suggesting that the contribution of external high-temperature and highsalinity fluids to the diagenetic fluid of Rd2 was not significant. The high temperature of the dolomitizing fluids of Rd2 may be influenced by ancient heat flow and geothermal gradient uplift [7], and in turn, the high geothermal gradient could favor the recrystallization of dolomite to form larger crystals of Rd2 [7]. Therefore, the dolomitizing fluid of Rd2 was primarily derived from the Permian seawater within the shallow-burial strata, mixing with a small portion of external hydrothermal fluids with high temperature, salinity, and 87 Sr/ 86 Sr ratio.
The Rd3 and Sd are primarily composed of high-temperature, high-salinity brines from the deep basin, and Permian seawater also contributed to their formation. The large and curved crystals observed in these samples indicate rapid crystal growth and high formation temperatures [41,43]. The δ 13 C values of the samples are consistent with the host limestone, suggesting that the carbon source is inherited.
The 87 Sr/ 86 Sr ratios of most Rd3 and Sd samples are higher than those of the contemporaneous Permian seawater and other types of dolomites, indicating the possible influence of either higher degrees of evaporation [57,58] or large-scale water-rock interaction with external fluids with high 87 Sr/ 86 Sr ratios [59,60]. However, the sedimentary environment in the study area during the Middle Permian period was a broad carbonate platform [61,62], and there is no evidence of evaporite minerals, indicating a lack of evaporation. Therefore, the most likely cause of the high 87 Sr/ 86 Sr ratios of the Rd3 and Sd is the second scenario, where the influx of a large volume of external fluids with high 87 Sr/ 86 Sr ratios leads to this feature. The increase in the 87 Sr/ 86 Sr ratio results from the infiltration of hydrothermal fluids enriched in 87 Sr from clastic rock strata of the basin into the diagenetic fluids (e.g., those from the lower Silurian and Ordovician strata; [12]), with the basement faults serving as the conduit for the fluids [45,63,64].
Moreover, the high homogenization temperatures and salinities of the fluid inclusions in Rd3 and Sd further support this conclusion (Figure 8). Additionally, the positive Eu anomalies also indicate that the diagenetic fluids in Rd3 and Sd experienced a deep reducing environment (Figures 9c,d and 10d) [65]. Compared to the host limestones, Rd3 and Sd samples exhibit a significant increase in Mn content (Figure 10c), which is the result of the input from external fluids. The Mn content variation is also reflected in the strong luminescence of Rd3 and Sd ( Figure 5) [66,67]. The Sr content of Rd3 is relatively lower than that of Rd2 because Sr is continuously lost with increasing dolomitization [68,69]. Furthermore, it is found that hydrothermal minerals, including fluorite and siderite, are filled within crystals of Rd3 and Sd (Figure 6d-f), indicating that the high-temperature and high-salinity fluid is hydrothermal fluid [1,9], as these mineral assemblages are products of hydrothermal activity.
Therefore, the high temperature, high salinity, and significantly positive Eu anomaly provide evidence that the fluid responsible for the dolomitization of Rd3 and Sd was mainly derived from hydrothermal fluids, which penetrated through the high-87 Sr-bearing clastic strata before entering the Middle Permian through deep faults. However, the presence of inherited carbon sources and the 87 Sr/ 86 Sr ratios of a small portion of the samples falling within the range of Permian seawater also suggest that residual seawater in the strata contributed to the formation of Rd3 and Sd.

Genetic Mechanism of the Dolomites
Based on the above analysis, the diagenetic fluids of the dolomites in the Middle Permian of the central Sichuan Basin are mainly derived from seawater and deep-seated hydrothermal fluids. The heterogeneity in the formation and spatial distribution of different types of dolomites resulted from the degree of involvement of different source fluids, hydrological circulation patterns, and differences in water-rock interactions in different stages of diagenesis. Considering the sedimentary characteristics, tectonic activity, burial characteristics, and paleotemperature characteristics of the central Sichuan Basin comprehensively, we propose that the genetic mechanism of the Middle Permian dolomites is the result of seawater-hydrothermal mixed dolomitization caused by ELIP activity.
Rd1 might be formed as the primary mineral along the calcite in the original seawater. The high mud content could facilitate the growth of primary dolomite through mechanisms of clay surface-induced dolomite formation. The low Mn in the limestone associated with Rd1 also indicates weak diagenesis. On the other hand, the conversion of aragonite to calcite in early seawater may be responsible for the higher Sr concentration in limestone ( Figure 10).
The genetic mechanism of Rd2 is interpreted as heated seawater dolomitization under shallow burial conditions. The dolomite seeds that emerged during the Rd1 phase are considered a critical constituent for the dolomitization process of Rd2. Additionally, the pre-and post-eruption periods of the ELIP led to a relatively high geothermal temperature that facilitated the extensive development of Rd2. The anomalous temperature rise was initiated by the thermal energy emanating from ELIP, which commenced at 290 Ma and terminated at 240 Ma [22]. Correspondingly, it has been confirmed that the formation time of the main types of Middle Permian dolomite in the central Sichuan Basin is mainly between 251 ± 11 Ma to 264 ± 10 Ma, which coincides with the time of ELIP activity, based on the existing U-Pb dating from the Maokou Formation in the southeastern Sichuan Basin [12].
The petrological and geochemical properties of the dolomite in the study area are similar to the aforementioned region of [12], and Rd2 is believed to be of shallow-burial and heated-seawater origin. Therefore, it is believed that the dolomitization in the study area was influenced by the ELIP activity. The activity of the ELIP mantle plume has caused the reactivation of a series of NE-SW basement faults in the central Sichuan Basin [30,31], providing pathways for the upward migration of deep hydrothermal fluids during the eruption period. Therefore, the Middle Permian strata in the central Sichuan Basin have been in a high paleotemperature environment for a long time, with a paleo-heat flow value as high as 130 mW/m 2 , due to the strong thermal effect of the ELIP [22,23].
As ELIP activity reached its peak, the intrusion of magma intensified deep-seated hydrothermal activity in the study area, and the exceptionally high thermal effect and pressure provided the driving force for the upward movement of deep hydrothermal fluids [70]. Deep faults and their associated fracture systems provided preferential pathways for the upward migration of hydrothermal fluids. High-temperature, high-salinity hydrothermal fluids rose to the shallow-buried Middle Permian strata and were released into the formation water (i.e., mainly remained coeval seawater) system. In addition, the extensive karstification caused by the uplift of the study area prior to the ELIP eruption formed many dissolution pores and fractures in the topmost Maokou Formation [26][27][28][29]. The basement faults and the dissolution caves and fractures provided pathways for communication between the deep hydrothermal fluids and the overlying seawater in the sedimentary strata, thus facilitating the mixing dolomitization of seawater and hydrothermal [7].
This seawater-hydrothermal mixed fluid circulated constantly in the high-permeability shoal facies of the Middle Permian strata, which were still shallowly buried. The continuous supply of seawater sources provided sufficient Mg 2+ , and the influx of hydrothermal fluids and the exceptionally high geothermal gradient provided high temperatures to break through the kinetic barrier of dolomitization [71]. Therefore, a large amount of Rd3 and Sd formed near the faults in this stage. Numerical simulation studies [72] have also confirmed that the mixing fluid moves, which is communicated through the basement faults, and can thus meet the Mg flux requirements for large-scale dolomitization and form dolomite along high-permeability fault systems. In addition, dolomitization has a selective nature in the vertical direction, and the mixed fluid of seawater and hydrothermal fluid preferentially migrates and circulates laterally along high-permeability strata near faults [7,73]. Therefore, the shallow buried Mao2 and upper-Mao3 strata deposited in intra-platform shoals were the preferred choice for injecting dolomitizing fluids and water-rock reaction, followed by the deep burial Qi2 strata. The relatively deeper buried condition led to a higher degree of rock compaction and lower permeability of Qi2 strata compared to that of the Mao2 and upper-Mao3 strata, resulting in less formation of dolomites in the Qi2 member.
However, the Qi2 member was relatively deeper buried and had a higher degree of rock compaction, resulting in lower permeability compared to the Mao2 member and relatively less seawater supply, and thus relatively less dolomite formation.
The petrological and geochemical evidence indicates that the formation of Rd2, Rd3, and Sd was continuous. With the continuous influx of deep hydrothermal fluids, the tectonic-hydrothermal Rd3 dolomite was formed around the fault system, while the seawater-dominated Rd2 dolomite was formed away from the fault. The hydrothermal fluid rose along the basal faults and was released as overpressure fluid, causing hydraulic fracturing on a large scale due to the fluid pressure being much greater than the rock pore fluid pressure [3,11,45,63,64,74,75]. Subsequently, the hydrothermal fluid deposited the Sd dolomite mainly in the fractured cracks and pores, forming zebra-like structures and hydraulic breccias.
The incorporation of deep hydrothermal fluids in the diagenetic fluids caused Rd3 and Sd samples to show more hydrothermal signals (e.g., abnormally high 87 Sr/ 86 Sr ratios, positive Eu anomalies, high-temperature, high-salinity features, etc.). Meanwhile, the hydrothermal acid dissolution could form large dissolution cavities inside Rd2 and Rd3 [76,77]. Then, Sd, cc, fluorite, siliceous, and other hydrothermal minerals precipitated, filled, and blocked the cavities and fractures. Thus, Rd3 and Sd are mainly distributed around the basement faults and are featured by limited lateral extension due to the massive hydrothermal fluids released from the deep faults.
Excessive recrystallization led to pore blockage, making it difficult for fluids to continue to flow laterally [78][79][80]. Once complete dolomitization and mineral stabilization have been achieved, even if more magnesium-rich fluids enter the system, the size of the dolomitization body will not increase [3,8,81,82], and Rd3 and Sd would remain concentrated in the faults and fracture networks. In contrast, as hydrothermal fluids migrated laterally in highly permeable beach facies, the proportion of hydrothermal fluids in the formation water system decreased, and more Rd2 was formed at the sites away from the hydrothermal source. This is also the reason why weak hydrothermal signals are reflected in Rd2 (e.g., 87 Sr/ 86 Sr ratios and salinity slightly higher than Permian seawater). If there were many dense companion faults in the basement fault system, the above dolomitization process could result in the connection of Rd2 largely and extend laterally along high-permeability shoal deposits. Therefore, we suggest a relatively continuous process of the formation of Rd2 to Sd dolomites. This is also supported by the U-Pb dating analysis that the formation of these types of dolomite occurred during the peak of ELIP activity [12].
In summary, the main cause of the formation of Middle Permian dolomites in the central Sichuan Basin is attributed to the seawater-hydrothermal mixing dolomitization via the effects of ELIP activity. Dolomitizing fluid circulation was facilitated by the activation of the basement faults, karst pores in the overlying rock, and anomalous high temperature from deep and shallowly buried high-permeability strata. During the peak of ELIP activity, continuous upwelling of deep hydrothermal fluids resulted in the persistent formation of Rd2, Rd3, and Sd. The dolomitizing fluids of Rd2 consisted mainly of seawater, making it the forefront of the dolomite body. The dolomitizing fluids of Rd3 and Sd, which were dominated by deep hydrothermal fluids, were primarily distributed along the fault system. Excessive dolomitization caused by hydrothermal activity might have inhibited the lateral migration of hydrothermal fluids.

1.
In the eastern part of the Middle Permian in the Sichuan Basin, dolomites mainly occur in the upper part of the Maokou Formation and the top of the Qixia Formation, consisting mainly of three types of replacive dolomites (Rd1, Rd2, and Rd3) and saddle dolomite cement (Sd).

2.
Petrological and geochemical evidence indicates that Rd1 formed during the initial stage of burial, mainly from residual seawater in the strata.

3.
Intense hydrothermal activity related to ELIP caused seawater-hydrothermal mixing dolomitization, which was the main cause of the formation of Rd2, Rd3, and Sd. Activated basement faults, freshwater dissolution cavities at the top of the strata, and high-permeability strata were the main pathways for fluid circulation. The diagenetic fluids of Rd3 and Sd contain more deep-seated hydrothermal fluids, reflected in their higher 87 Sr/ 86 Sr ratios, distinct Eu-positive anomalies, and very high temperature and salinity. Their distribution is also controlled by basement faults.

4.
This study provides a good example for investigating the formation mechanism of dolomite controlled by faults under the influence of anomalous thermal effects.

Data Availability Statement:
No new data were created or analyzed in this study. Data sharing is not applicable to this article.